In 1988, a National Dropout Statistics Task Force (comprised of representatives from 27 states, 3 U.S. territories, and the District of Columbia) was created to work with the National Center for Education Statistics (NCES) to develop procedures for uniform counting of dropouts and dropout rate reporting. A standard definition was constructed for "dropout," and three alternatives for dropout rate definitions were considered, dividing full-year dropout counts by: (1) initial fall membership; (2) spring membership plus fall-to-spring dropouts; and (3) following fall membership plus fall-to-fall dropouts. A field test involving 26 states, the District of Columbia, and 2 outlying areas was conducted to determine whether these definitions could be consistently and uniformly applied. Data submitted during fall 1989, spring 1990, and fall 1990 provided memberships by race, ethnic category, and sex for students in grades 7 through 12. All participating areas (ranging from a high of 232 to a low of 223 districts) responded to surveys about practices affecting dropout rates. Data for 187 districts supported the analyses. Focus was on data for 456 transfer and 192 dropout cases. Only negligible differences were found among the three alternatives. Data show a clear increase in dropout rates from about 1% in grade 7 to about 6% in grades 10, 11, and 12, with a typical synthetic cohort dropout rate for grades 7 through 12 of 23%. Accuracy of dropout rates was generally good, and states agreed on the necessity of consistent reporting and valid dropout statistics. Twelve figures and 28 tables are provided. Seven appendixes present the surveys used, site visit guides, a school leaver status validation form, a tracker's guide, and an additional six tables. (SLD)